The study was designed in response to the NIMH RFA inviting research to "test models of depression recognition and treatment to prevent and reduce suicidal behavior in older patients in primary care settings". Elderly suicide most frequently occurs in the context of mild to moderate depression. In primary care patients, suicidal ideation is a risk factor for suicide and has been identified almost exclusively in patients with depressive symptoms and signs. Since most suicide victims are seen by their primary care physicians within a few weeks of their death, intervening at the doctor's office may be life saving. This study will investigate the effectiveness of an intervention aimed at improving the recognition of suicidal ideation and depression by the practices and facilitating the implementation of a treatment algorithm based on the AHCP guidelines. The linchpin of the intervention is the use of Depression Specialists (DS) who will collaborate with the physicians and help them to increase recognition of depression, offer timely and appropriately targeted treatment recommendations and encourage patients to adhere to treatment. In addition, procedures aimed at educating patients, families and physicians on depression and suicidal ideation will be implemented. We expect that this approach will lead to a reduction of depressive symptomatology and suicidal ideation and behavior and generate a practice model that has the ability to incorporate the advances of our clinical science. The study will be conducted by the Intervention Research Centers (IRC) of Cornell, University of Pennsylvania, and University of Pittsburgh. The intervention will be offered in 6 primary care practices from 3 geographic areas (metropolitan and suburban New York, Philadelphia, and Pittsburgh) and its impact will be contrasted to that of 6 comparable practices offering usual care. Data will come from a representative sample of patients aged 60 years an older selected through an age-stratified, two-stage sampling design and followed for two years. The group will consist of 1,200 subjects and comprise patients with depressive symptoms and signs (CES-D(11; estimated total N=920; per Center N=307) and a random sample of patients without significant depressive symptomatology (CES-D(11;estimated total N=280; per Center N=93). Beyond direct systematic clinical assessment of patients, information will be obtained on health services utilization from practice-based medical records an on cause of death from death certificates. Data will also be collected to document the impact of intervention on patient care, and on physician knowledge, attitudes, and satisfaction and test hypotheses derived from preliminary studies of the three IRCs.